mbaddeley/contikipy

Automated cooja simulations, log parsing, and plotting with python.

ContikiPy - Automated Cooja Simulations, Log Parsing, and Plotting with Python.

Intro

This repo hosts the source code for ContikiPy, a suite of Python scripts for automated simulation running in cooja, as well as log parsing and plotting.

About

ContikiPy is a pet project that began as an attempt to try and make the task of getting information and results out of the Contiki logs a bit easier. There has been no formal design as such, just a bunch of things that seemed like a good idea to try and do.

Current YAML configurations are set up for my own experiments (based on Contiki-NG logging), but they should provide a good template for making your own.

Development

I work with Contiki a fair amount so I'm constantly trying to improve usability, but if you have a pull request / any suggestions then feel free to contribute!

Getting Started

Running ContikiPy

You can run ContikiPy through the following command, where you replace "my-config-file-xx" with your own YAML configuration:

• ./contikipy.py --conf="my-config-file-xx.yaml" --runcooja=1 --parse=1 --comp=1

A full list of commands is available through:

• ./contikipy.py --help

Options

The three main options are as follows:

• runcooja - run simulations according to the YAML config
• parse - parse simulation logs according to the YAML config and output plots
• comp - compare plots according to the YAML config

Adding New Plots

1. In the yaml config:

• Add the name of your plot to 'plot:' in your yaml simulation.

2. In cplogparser.py:

• Add your plot to 'atomic_function_map' in plot_data() in cplogparser.py.
• Map your plot name to the plot function name.
• Add your plot to 'atomic_dict_map' in plot_data in cplogparser.
• Map your plot name to the required data sets.
• Write your plot function. It will take in a list of pandas dataframes (that you have previously mapped to this function) which you can use for generating your plot.